Psychological Science Compassion Training Alters Altruism and Neural Responses to Suffering
Helen Y. Weng, Andrew S. Fox, Alexander J. Shackman, Diane E. Stodola, Jessica Z. K. Caldwell, Matthew C. Olson, Gregory M. Rogers and Richard J. Davidson Psychological Science published online 21 May 2013 DOI: 10.1177/0956797612469537 The online version of this article can be found at:

Published by:

On behalf of:

Association for Psychological Science

Additional services and information for Psychological Science can be found at: Email Alerts: Subscriptions: Reprints: Permissions:

>> OnlineFirst Version of Record - May 21, 2013 What is This?

Downloaded from at Harvard Libraries on May 22, 2013

Caldwell1. Studies indicate Corresponding Author: Richard J. 2010. Despite the clear societal benefits of cultivating compassion. and difficult persons. Revision accepted 11/1/12 Compassion and altruism are of great interest to philosophical and scientific inquiry because of their central role in successful societies (Darwin. including loved ones. strangers. Eisenberg. Darwin. In healthy adults. 1871/2004. Goetz.1177/0956797612469537 Research Article Compassion Training Alters Altruism and Neural Responses to Suffering Helen Y. These results suggest that compassion can be cultivated with training and that greater altruistic behavior may emerge from increased engagement of neural systems implicated in understanding the suffering of other people. Johnstone. 1759/2010. and in DLPFC connectivity with the nucleus accumbens. University of Wisconsin–Madison. & Simon-Thomas.. Gregory M.5. including the inferior parietal cortex and dorsolateral prefrontal cortex (DLPFC).Compassion Training and Altruism Psychological Science OnlineFirst.4. Fox1. 5Department of Psychiatry. Fehr & Fischbacher.sagepub. University of Wisconsin–Madison. 2003. promoting intimate bonds between partners. Furthermore. WI 53705-2274 E-mail: rjdavids@wisc. Sober. & Spinrad. Stodola2. University of Wisconsin–Madison. BrefczynskiLewis. Weng1. emotional control. Keltner. and facilitating cooperation among genetically unrelated strangers ( at Harvard Libraries on May 22..4. Fabes. & Davidson. Diane E.3. Davidson1. we found that compassion training increased altruistic redistribution of funds to a victim encountered outside of the training context. Matthew C. Davidson. Goetz et al. compassion is cultivated toward different people.3. little is known about whether compassion and altruism can be trained and about the neural mechanisms that might underlie such effects. fMRI. Jessica Z.sagepub. Madison. Andrew S.2. 3Center for Investigating Healthy Minds at the Waisman Center.6. 2013 as doi:10. Such acts include enhancing the welfare of vulnerable offspring. and Richard J. 2008). social behavior.7. Compassion is the emotional response of caring for and wanting to help those who are suffering (Batson. Sober et al.2. & Wilson. 1997). 4HealthEmotions Research Institute..3. Rogers5. We examined whether compassion may be systematically trained by testing whether (a) short-term compassion training increases altruistic behavior and (b) individual differences in altruism are associated with training-induced changes in neural responses to suffering.nav DOI: 10. In compassion training. 1999). and reward processing. published on May 21.2.. University of Wisconsin– Madison. neuroimaging. Sober et Department of Psychology. 1999).. 2006. Alexander J. K. Shackman4. 1999). but little is known about individuals’ capacity to cultivate compassion through training. individual differences Received 3/21/12.2. Olson2. emotion regulation. University of Wisconsin–Madison. 2010) and may have evolved in humans to foster altruistic acts that increase survival of kin as well as nonkin (Darwin. Brown University Abstract Compassion is a key motivator of altruistic behavior. 2010. Brown University.1177/0956797612469537 pss. 1871/2004. 1991. meditation. increased altruistic behavior after compassion training was associated with altered activation in brain regions implicated in social cognition and emotion regulation. 2Waisman Laboratory for Brain Imaging and Behavior. and 7Miriam Hospital. 1500 Highland Ave.5 1 Psychological Science XX(X) 1­–10 © The Author(s) 2013 Reprints and permissions: sagepub. Wilson. as well as toward the self (Salzberg.1177/0956797612469537Weng et al. executive and emotional control. Smith. Waisman Laboratory for Brain Imaging and Behavior. Keywords compassion. altruism. decision making. 1871/2004.469537 research-article2013 PSSXXX10. Goetz et Center for Investigating Healthy Minds. 1991. Contemplative traditions claim that compassion can be enhanced with meditation training and that this results in greater real-world altruistic behavior (Lutz. 6Department of Psychiatry and Human Downloaded from pss. University of Wisconsin–Madison. 2013 .

Wager. Desbordes. 2008). Pek.g. given its role in controlled processing (Miller & Cohen. Silani. et al. positive emotions toward people who are suffering (Klimecki. we hypothesized that greater prosocial behavior after compassion training would be associated with higher levels of activation in anterior insula. We hypothesized that compassion training would increase altruistic behavior by enhancing neural systems involved in (a) the recognition and understanding of another person’s suffering and (b) emotion regulation of Weng et al. If the neural representation of suffering is increased by compassion training. 2007. & Raison. which reduces negative arousal and avoidance behavior. positive affect (Fredrickson. 2008). 2010. and reappraisal trainees practiced reinterpreting personally stressful events to decrease negative affect. Urry et al. 2001). Reappraisal training provided an ideal control for compassion training because although the combination of decreased distress and increased empathic concern predicts helping behavior (Batson. 2013). Tenzin Negi. whereas the goal of reappraisal training was to decrease one’s personal distress (Ochsner & Gross. We specifically tested these hypotheses against the reappraisal group. 2005). Expert meditation practitioners show greater empathic neural responses when listening to sounds of other people’s suffering during compassion meditation practice than control subjects do (Lutz. et al.. & Damasio. Mayr. 1991. 2012). & Buckner.. Compassion training also enhances responses toward other people. Zaki & Ochsner. This greater prosocial behavior would also be correlated with increased activation in nucleus accumbens (NAcc). which strengthens the motivation to approach and relieve another person’s suffering (Batson. in which the psychological goal was self-focused (to decrease one’s own suffering) rather than other-focused (to decrease other people’s suffering through compassion). The neuroscience of empathy highlights two systems for understanding the states of other people: experience sharing. & Singer. Vincent. Coffey.. Singer et al. & Eisenberger. Raichle.sagepub. Miller. Further. Brefczynski-Lewis. Compassion trainees cultivated compassion for different targets. we measured brain activation before and after 2 weeks of training using functional MRI (fMRI) and investigated whether increased altruism could be explained by training-induced changes in the neural response to human suffering.. & DeSteno. responses to suffering that support affiliation and helping behavior. 2008). & Burghart. & Gross. 2011. 2008.. which has been implicated in studies of empathy and compassion (Immordino-Yang. and fronto-parietal control networks (Dosenbach. & Singer.. 2008). given the amygdala’s role in responding to negative stimuli and distress (Zald. 2006). Snyder. 2008). Lamm et al. et al. 2006) and positive appraisals of aversive stimuli (Wager et al. Furthermore. Lamm. Recent work suggests that compassion training can increase prosocial behavior (Leiberg. 2009). To rigorously test these hypotheses. which has been linked to charitable giving (Harbaugh. 2008. Schlaggar. we predicted that greater altruism in compassion trainees would be associated with increased activation in prefrontal cortex (PFC).. BrefczynskiLewis. The neural mechanisms by which compassion training alters altruistic responses to suffering remain unknown. or (b) increasing empathic concern. Fair. 1991). which involves explicitly considering and understanding others’ mental states through social inferences as well as through selfreferential processes (Lamm. but they differed in that the goal of compassion training was to increase empathic concern and the desire to relieve suffering (Lutz. & Finkel. Masten. Decety. 2008). we investigated whether short-term compassion training would enhance altruistic behavior toward a victim encountered outside of the training context. Morelli. 2008).. and mentalizing.. and psychological and physical health (Fredrickson et al. & Singer. 2003). Kahn. Although many of the same regions are implicated in reappraisal as in compassion (Wager et al. Batson. Damasio. at Harvard Libraries on May 22. & Ochsner. Seppala.. Rilling. We also predicted that compassion training would be associated with decreased amygdala activation. 2006) and predicts helping behavior (Hein. 2006. which involves vicariously sharing the states of others. Klimecki. Condon. Brefczynski-Lewis. then regulatory systems are needed to respond to this suffering with an approach rather than an avoidance response.. Lutz. 2012). emotion regulation (Ochsner & Gross. In response to suffering. Prior theoretical and empirical work suggests that altruistic responses toward another person’s suffering can be strengthened through either of two regulatory pathways (Decety & Jackson. 2008). 2009. reappraisal training only decreases distress without enhancing concern. Hutcherson. Preuschoff. increases in PFC activity) would not be associated Downloaded from pss. 2011. 2008. Cohn. In the study reported here. 2011.2 that compassion training can improve personal wellbeing. including stress-related immune responses (Pace et al. McColl. & Petersen. Altruistic behavior was assessed using the redistribution game. we compared altruistic responses of participants given compassion training with responses of participants given an active control intervention of reappraisal training. in press). 2013 . Lindquist. 2006): (a) decreasing personal distress. Both interventions trained emotion-regulation strategies that promote well-being. Davidson. a novel economic decisionmaking task that models both unfair treatment of a victim and costly redistribution of funds to the victim. Moll et al.. 2011). and empathic accuracy (Mascaro. Hughes. we expected that the hypothesized changes (e. 2005. Cohen. Eisenberg et al. & Singer.

and the final sample consisted of 41 participants who believed that they were interacting with real players in the redistribution game (the other 15 participants expressed suspicion about the manipulation and were therefore excluded from the analysis. At the end of the entire protocol.) Participants were told that they were playing the game with live players over the Internet. we scanned participants using fMRI before and after training while they employed their assigned emotion-regulation strategy. 2003).1 Altruistic behavior task: redistribution game. participants could choose to spend any amount of their own endowment ($5) to compel the dictator to give twice that amount to the victim (Fig. and a difficult person). 8 male. following which he or she practiced the fMRI task in a mock MRI scanner.  To determine whether altruistic behavior was predicted by changes in neural responses to human suffering. during this visit. Using anonymous online interactions. This economic decision-making task models both unfair treatment of a victim and costly redistribution of funds to the victim.  Training consisted of practicing either compassion or reappraisal using guided audio instructions (via the Internet or compact disc) for 30 min per day for 2 weeks. The groups did not differ in age. right-handed. Method Participants Fifty-six participants completed the entire protocol. or the amount of practice time they received. 8 male. 1a). never instructing participants to use the training they received. describing it in purely economic terms. and reappraisal trainees practiced reinterpreting personally stressful events to decrease negative affect (see Trainings and Fig. Visit 2 occurred approximately 1 week later. Procedure Overview. 13 female. reappraisal trainees were instructed to decrease negative emotions by silently reinterpreting the emotional meaning of the images. For more details about the procedure. feelings of compassion for different targets (a loved one. baseline trait compassion. After witnessing this violation of the fairness norm (Fehr & Fischbacher. Compassion trainees were instructed to evoke feelings of compassion while silently repeating compassion-generating phrases. fMRI task and stimuli. No participant had issues that would pose a risk for his or her safety in the scanner.9 years) or reappraisal training (n = 21. Participants were paid the amount that was left in their endowment after making the decision. Participants were healthy adults (18–45 years of age). Each participant was randomly assigned to receive either compassion training (n = 20. participants first observed a dictator (endowed with $10) transfer an unfair amount of money ($1) to a victim who had no money (Fig. Visit 3 occurred immediately after the 2 weeks of training were completed. We tested whether compassion training could affect altruistic behavior outside of the training context using the redistribution game. and enforcing real monetary consequences for participants’ behavior. 2013 . all participants observed the same preprogrammed unfair dictator offer.  Participants came to the laboratory on three occasions. mean age = 21. (See Supplementary Method and Analyses in the Supplemental Material for fMRI data-acquisition parameters). participants completed the pretraining fMRI scan and began training later that day. Because compassionate behavior is specifically evoked by unfairness. 1b). In contrast. a stranger. and had no previous experience in meditation or cognitive-behavioral at Harvard Libraries on May 22. Trainings. and attended an fMRI session both before the start of training and after training finished. All subjects gave informed consent and were paid for participation. see Supplementary Method and Analyses in the Supplemental Material. participants were debriefed and asked whether they believed they were playing against real people in the game.5 years). Data were analyzed only for participants who believed the paradigm (see Table S2 in the Supplemental Material). Participants in the two groups were presented with images of human suffering and nonsuffering (neutral condition). Compassion trainees practiced cultivating Downloaded from pss. The experiment was approved by the University of Wisconsin–Madison Health Sciences Institutional Review Board. this visit included the posttraining fMRI scan and the altruistic behavior task (performed outside of the scanner). Effects of demand characteristics on behavior were minimized by presenting the game as a unique study.sagepub.Compassion Training and Altruism 3 with altruistic behavior because the behavior is not congruent with reappraisal’s goals. removing the physical presence of players and experimenters during game play. mean age = 22. S1 in the Supplemental Material). (See Supplementary Method and Analyses in the Supplemental Material for full details of the redistribution game. completed 2 weeks of training (11 out of 14 practice days were required). gender. see Tables S1 and S2 and Supplementary Method and Analyses in the Supplemental Material available online for further information about the sample). At Visit 1. 12 female. the self. each participant was randomly assigned to compassion training or reappraisal training and briefly instructed in the assigned strategy.

com at Harvard Libraries on May 22. Because of these violations of normality. a Dictator $10 – $1 $1 Victim $0 + $1 b Dictator $10 – $1 – (X*2) (X *2) Victim $0 + $1 + (X*2) Participant $5 Participant $5 – X c Redistribution (rank) 30 25 20 15 10 * Compassion Training Reappraisal Training Fig.g. the dictator (endowed with $10) transfers an unfair amount of money ($1) to the victim while the participant (endowed with $5) observes.05). physical pain. Images were balanced across sets for published normative ratings of valence and arousal. such as working or walking down a street.1).4 Weng et al. randomized).5. In the second stage (b). Behavioral analysis Across all participants. 7 = most arousing). The graph (c) shows the average rank-transformed redistribution amount as a function of the type of training participants received. participants saw each image again for 2 s and rated the arousal of each image (1 = least arousing. In the first stage of the redistribution game (a). Images of suffering depicted emotional distress. Parametric tests were then Downloaded from pss.sagepub. They then applied the assigned regulation strategy to a series of 12 images. SE = 0. or acts of violence (e. Error bars denote standard errors of the mean. a crying child). X = $2..5 of 41) results in an equal distribution between the dictator and the victim ($5 each).  Paradigm of and results for the redistribution game. and saturation (all ps > . Redistribution of $4 (i. rank = 35. and 2 participants qualified as outliers (> 3 SD from the population mean). The asterisk indicates that there was a significant difference between groups (p < . Blocks ended with a final fixation baseline (17–38 s). the redistribution response was positively skewed (skewness = 1. Image randomization was performed once for each set and then fixed. the participant can spend any amount (X) up to $5 to compel the dictator to give twice that amount to the victim. as well as for properties of hue. Participants then received both an auditory and visual instruction (3 s) stating that they should invoke either “compassion” or “reappraisal” (depending on group assignment). we rankordered the behavioral response across both groups so that strong assumptions were not made about the scaling or normality of the residuals.e.. Participants were instructed to regulate their emotional responses to the images over three blocks. Images were pseudorandomized so that three or more images from either condition were not presented in a row. Two parallel sets of images (20 suffering and 16 neutral) were created to ensure that participants viewed different images before and after training. Each image was presented for 12 s and separated by a fixation interval (5–11 s. After each block. luminance.37). Neutral images depicted people in nonemotional situations. Set order was counterbalanced and randomized. 1. 2013 . which was followed by a fixation cross (5–7 s). Each block began with a 20-s fixation baseline period. a burn victim.

First.01 after an initial voxel-wise threshold of p < . data from the Group × Redistribution Rank interaction were Downloaded from pss. insula..sagepub. Voxel-wise regression maps were corrected for multiple comparisons (p < .01 after a conjunction voxel-wise threshold of p < . whole-brain analyses were conducted and corrected for multiple comparisons (p < . and NAcc). we performed a PPI analysis using the DLPFC seed region identified by the IPC conjunction analysis. and beta coefficients were converted to percentage signal change (PSC). IPC conjunction analysis. These values were used for descriptive and diagnostic purposes only (Vul.01 after an initial voxel-wise threshold of p < . For in-depth analyses and discussion of redistribution values and ranks. we performed an independentsamples t test on the ranks. This analysis identified the right inferior parietal cortex (IPC). we performed a second-level Group × Redistribution Rank voxel-wise analysis. & Pashler. Training-induced PPI changes were calculated by subtracting PPI betas before each scan from PPI betas after each scan.01). mean PPI-change betas were extracted from the clusters for each participant and analyzed to yield parameter estimates and determine the directionality of the relationship for each group. 2013 . a second-level Group × Redistribution Rank voxel-wise analysis was performed. controlling for main effects of group and redistribution. whole-brain corrected at p < . Winkielman. Finally. we investigated whether reported arousal was associated with either redistribution or neural changes. and NAcc. In region-of-interest (ROI) analyses. Interaction analysis. This analysis identified a cluster in the dorsolateral PFC (DLPFC. Correlational analyses. 1991. Psychophysiological interaction (PPI) analysis. insula. 2009). after training) using standard first-level analyses (see Supplementary Method and Analyses in the Supplemental Material). The PPI regressor consisted of comparing DLPFC connectivity in response to images of suffering with DLPFC connectivity in response to neutral images. Each participant’s neural response to suffering during regulation was estimated with the contrast between activation to images of suffering and activation to neutral images at each fMRI scan time point (before training. See Supplementary Method and Analyses in the Supplemental Material for full details. corrected for multiple comparisons (p < . To decompose the interactions. To determine regions in which altered PFC connectivity predicted higher amounts of redistribution in compassion trainees than in reappraisal trainees. To identify regions where training-related changes specifically predicted greater redistribution in compassion trainees than in reappraisal trainees. see Supplementary Method and Analyses in the Supplemental Material. we performed a conjunction analysis requiring voxels to be (a) correlated with changes in IPC activation across participants in both groups (voxel-wise p < . To test this. To identify regions where trainingrelated PPI changes specifically predicted greater redistribution in compassion trainees than in reappraisal trainees.01) using Monte Carlo simulation. A series of tests were conducted to identify regions in which changes due to training predicted greater redistribution in compassion training than in reappraisal training. A whole-brain interaction contrast (Group × Redistribution Rank) was tested on neuralchange scores. For descriptive purposes. Training-induced changes were calculated by subtracting PSC values before each scan from PSC values after each scan. controlling for main effects of group and redistribution. Eisenberg et al. we computed arousal-change scores (analogous to the neural-change scores) and correlated them with altruistic redistribution in each group. fMRI analyses Overview.and within-subjects analyses to identify regions that were functionally connected to clusters identified in the interaction analysis and networks involved in emotion regulation.01 after an initial voxel-wise threshold of p < .  Compassion training may increase altruistic behavior by decreasing personal distress evoked by suffering (Batson. 2006).Compassion Training and Altruism 5 performed on the ranked data. To examine whether changes in arousal were associated with changes in neural responses to suffering.01) using Monte Carlo simulation within each bilateral ROI (amygdala.01) using Monte Carlo simulation within bilateral a priori ROIs of the amygdala. To identify regions that may be functionally connected to the IPC in order to increase the amount participants redistributed in each training group.001). we computed correlations between arousal-change scores and neural-change scores identified in the previous fMRI analyses in each group. at Harvard Libraries on May 22. To test the mean difference between groups.  Individual functional and structural MRI brain data were preprocessed and normalized to Montreal Neurological Institute (MNI) space.01) and (b) identified in the original Group × Redistribution Rank interaction (voxel-wise p < . Follow-up tests were conducted using both across. we extracted mean PSC-change scores from the clusters for each participant and analyzed them to yield parameter estimates and determine the directionality of the relationship for each group.

corrected within the ROI. Lamm et al. reappraisal training: r(19) = .05. 2008). even when individuals are not explicitly cued to generate compassion. corrected. p < . p < . independent-samples t(39) = 2. The whole-brain Group × Redistribution Rank interaction test revealed that training-induced changes in right IPC activation were differentially associated with altruistic redistribution in the two training groups (Fig. r(19) = −. but not in reappraisal trainees. S4 and Supplementary Method and Analyses in the Supplemental Material for exploratory analyses within the ROIs.  Emotion regulation is thought to involve the influence of the PFC over other regions such as the amygdala. and both regions differentially predicted redistribution between groups.70.92. 2013 . $0. DLPFC were highly coupled—compassion training: r(18) = . d = see Tables S3 and S4 in the Supplemental Material).sagepub. We found that decreases in reported arousal to images of suffering were correlated with increased redistribution in compassion trainees. Within the a priori ROIs. S3 in the Supplemental Material). 2c. The changes in IPC and Discussion Individuals who trained in compassion for 2 weeks were more altruistic toward a victim after witnessing an unfair social interaction compared with individuals who trained Downloaded from pss. p < . These findings suggest that frontoparietal executive control networks (Dosenbach et al. 1c). no region survived correction at p < . also see Table S5 in the Supplemental Material). p < .01 whole-brain corrected). reappraisal trainees increased the distribution by only 31%. S2 in the Supplemental Material). 1991. Compassion trainees who showed greater DLPFC-NAcc connectivity redistributed more funds after training. DLPFC PPI connectivity changes and altruistic redistribution.. we found a significant interaction in the NAcc. In contrast. greater IPC activation due to training was associated with greater redistribution. 2008. In compassion trainees. Compassion trainees also spent more than individuals with no training in the validation sample (Fig.59. we tested whether changes in task-related functional connectivity between DLPFC and amygdala. whereas reappraisal trainees who showed greater DLPFC-NAcc connectivity redistributed less money after training (Fig. Arousal correlations with altruistic redistribution and neural change. or DLPFC and insula predicted greater altruistic redistribution in compassion training than in reappraisal training..84 times more money than reappraisal trainees ($1. 3c. Keysers. 2008) may be recruited by compassion training in order to regulate emotions and increase altruistic behavior.. 3a). demonstrating that DLPFC-NAcc connectivity was differentially associated with redistribution in compassion training compared with reappraisal training (Fig. 2004. r(19) = .21). compassion trainees spent more money to redistribute funds to the victim compared with reappraisal trainees (Fig. 2d. r(18) = −.6 Weng et al.. No relationship was found in the insula or the amygdala. see also Table S5 in the Supplemental at Harvard Libraries on May 22. In the main study. 1980) redistributed more money. Urry et al. Results Altruistic redistribution Findings in an independent validation sample2 (N = 72) confirmed that altruistic redistribution is a behavioral signature of compassion: Individuals who endorsed greater levels of trait empathic concern (Davis. DLPFC and NAcc. Using the DLPFC cluster defined by the IPC conjunction test as a seed (Fig. The IPC conjunction test identified only the DLPFC (Fig. after 2 weeks of training. We hypothesized that greater altruism resulting from compassion training would be predicted by training-related changes in the neural responses to images of suffering.45. The IPC is implicated in experience sharing as part of the mirror-neuron network (Gallese. where greater increases in DLPFC activation predicted greater altruistic redistribution in compassion trainees. see Tables S3 and S4 in the Supplemental Material).  Compassion training may increase altruistic behavior by decreasing personal distress evoked by suffering (Batson. 3b. p = . This demonstrates that purely mental training in compassion can result in observable altruistic changes toward a victim. but this was not found in reappraisal training (Fig. 2b. also see Table S5 in the Supplemental Material).09. r(70) = . 2006). and the opposite relationship was found in reappraisal trainees (Fig. 2005. 2a. also see Tables S3 and S4 in the Supplemental Material. Neuroimaging Group differences in neural change and altruistic redistribution. see Supplementary Method and Analyses in the Supplemental Material for discussion of the directionality of the connectivity). p < . p < . p = . Wager et al.. Compassion trainees spent 1. We further investigated whether decreases in arousal were associated with neural changes and found that greater DLPFC-NAcc connectivity was correlated with decreases in arousal in compassion trainees.79. Decreases in arousal were not associated with IPC or DLPFC changes in either group (all ps ≥ . Eisenberg et al.64. and we investigated whether the IPC was functionally connected to other regions that also differentially predicted redistribution between groups. insula. but not in reappraisal trainees.43.14 vs. p < . respectively) and increased the distribution between the dictator and the victim by 57%. and NAcc (Ochsner & Gross.001. & Rizzolatti. p < .001 (Fig. Vincent et al. p < . 2006. Using PPI.65.05. See Fig.13. r(18) = −. 2011)..

and up-regulation of positive emotion systems. This demonstrates that a purely mental training can generalize to behavioral domains by affecting social behavior outside of the training context.5 ∆ BOLD (% Change) ∆ BOLD (% Change) Fig.05.  Activation in right inferior parietal cortex (IPC) and right dorsolateral prefrontal cortex (DLPFC). Greater IPC activation specifically predicted greater redistribution in compassion trainees and not in reappraisal trainees. ***p < ..5 of 41).29*) Redistribution (rank) Redistribution (rank) 40 30 20 10 0 –1.23) d 50 R DLPFC Compassion Training (semipartial r = . which provides evidence for functional neuroplasticity in the circuitry underlying compassion and altruism. The blue outline in (c) indicates that the cluster was identified by a conjunction test. Images are in Montreal Neurological Institute space.18 R R b 50 R IPC Compassion Training (semipartial r = . 2004) and may reflect increased simulation of the suffering of other people. which results in an equal $5 distribution between the dictator and the victim.Compassion Training and Altruism 7 a c Interaction R 2 . increases in altruistic responses were correlated with training-related changes in the neural response to suffering. from before to after training in compassion trainees.sagepub. The pattern of neural changes in compassion training suggests that increased altruistic behavior is achieved by enhancing neural mechanisms that support the understanding of others’ states.5 0 0. This region has been implicated in the human mirrorneuron system (Gallese et at Harvard Libraries on May 22. 2. in both sessions. The dashed line indicates the rank of a redistribution of $4 (rank = 35. 2013 . respectively. If Downloaded from pss.001). Furthermore. The color coding indicates the amount of variance accounted for by the Group × Redistribution Rank interaction. which suggests that IPC recruitment is a unique neural marker for altruism induced by compassion training.5 –1 –0.5 Equal Distribution –1 –0. in reappraisal and individuals in a validation control group.39 .49***) Reappraisal Training (semipartial r = –. participants regulated their emotional responses while viewing images of human suffering. The brain images in (a) and (c) show bloodoxygen-level-dependent (BOLD) changes in right IPC and right DLPFC.5 Equal Distribution 40 30 20 10 0 –1. Asterisks indicate significant results (*p < . R = right.5 0 0.49***) Reappraisal Training (semipartial r = –. greater fronto-parietal executive control. The scatter plots (with best-fitting regression lines) in (b) and (d) show rank-transformed redistribution amounts for the two training groups as a function of the percentage signal change in BOLD responses from before training to after training.

2005) of helping the victim. This may represent increasing positive appraisals of aversive stimuli (Wager et al.. 2001) to help other individuals. Asterisks indicate significant results (**p < . Vincent et al.  Results for connectivity between right dorsolateral prefrontal cortex (DLPFC) and nucleus accumbens (NAcc).. The image in (b) shows the regions of NAcc in which there was differential activation in response to images of human suffering before and after training in compassion trainees. decreased reported arousal after compassion training may be due to enhancement of reward-related neural systems. Vincent et al.41**) Redistribution (rank) 40 30 20 10 0 –30 Equal Distribution R –20 –10 0 10 20 30 b Connectivity (β) y=8 Interaction R 2 .g. ROI = region of interest. as reflected by the increased DLPFC-NAcc connectivity that predicted redistribution in compassion trainees. 3. 2013 .com at Harvard Libraries on May 22. Furthermore. a c 50 R DLPFC-NAcc Compassion Training (semipartial r = . the signal of other people’s suffering is indeed increased by compassion training.5 of 41). Regulation of internal information may include increasing positive emotions toward other people’s suffering. this leads to an emotion-regulatory challenge that requires trainees to approach rather than avoid suffering in order to engage in prosocial behavior. The scatter plot (with best-fitting regression lines) in (c) shows rank-transformed redistribution amounts for the two training groups as a function of changes in right DLPFC-NAcc connectivity.36 . Images and coordinates are in Montreal Neurological Institute space..sagepub.8 Weng et al.. 2008) by enhancing the reward value of the victim’s well-being (e. which results in an equal $5 distribution between the dictator and the victim. R = right.07).27†) Reappraisal Training (semipartial r = –. These findings Downloaded from pss.17 ROI (NAcc) R Fig.01.. The coordinated activation of the IPC and DLPFC in compassion trainees may reflect greater sustained attention and goal maintenance (Miller & Cohen. 2008). The color coding indicates the amount of variance accounted for by the Group × Redistribution Rank interaction. †p = . 2008. as well as integration of information from systems that process both external information (of other people’s suffering) and internal information (the goal to help. The dashed line indicates the rank of a redistribution of $4 (rank = 35. This transformation of emotional response may have been instantiated by a fronto-parietal executive control network (Dosenbach et al. which was used as the seed region in the psychophysiological interaction analysis. caring) and increasing the anticipated reward (Knutson & Cooper. 2008) in order to increase altruistic behavior in compassion trainees. The image in (a) shows the DLPFC cluster identified by the conjunction test.

689–695.. Nature. The amygdala and ventromedial prefrontal cortex in morality and psychopathy. (1980). S. Supplemental Material Additional supporting information may be found at http://pss . A. Lerner (Eds. & Jackson. emotional valence and arousal may be measured using methodology that is less susceptible to demand characteristics. such as facial electromyography and skin conductance response. A clear limitation of this study is that altruistic behavior was not measured at pretraining. Davidson). Fair. neural changes may have resulted in decreased helping of other individuals in order to serve the primary goal of decreasing personal negative affect. Y. Seltzer. References Batson.. Open hearts build lives: Positive emotions. Grossman. M. C. England: Penguin. self-focused). Social influences on neuroplasticity: Stress and interventions to promote wellbeing. NY: Wiley. Decety. W. D. & DeSteno. M. A social-neuroscience perspective on empathy. medicine) as well as clinical subgroups (Hofmann. Declaration of Conflicting Interests The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article. Kluge Foundation. Current Directions in Psychological Science. Pek. longitudinal designs should be employed to determine the length of compassion training needed to have sustained behavioral effects. R.. In sum. J. Fabes. the goal was to increase caring for people who are suffering and to help. In reappraisal training. J. (2008). A. 2007). (1991).. L. Training audio files and written scripts can be downloaded at www. 2012). and gifts from Bryant Wangard. A. S. 15. The nature of human altruism. R. 2011). Fox and H. 425. the Mind and Life Institute (H. & Petersen. S. Blair. E. which may strengthen claims that compassion training increases altruism (Leiberg et al. (2007). principal investigator). John Templeton Foundation. D. 2010. induced through loving-kindness meditation. Damon. A. Gallese. The descent of man and selection in relation to sex. greater changes in these regions predicted less redistribution in reappraisal trainees. J. & Davidson. and Ann Down (R. 85. 2011) characterized by deficits in compassion. J. Davidson. Fredrickson. 99–105. Handbook of child psychology. and this may explain the crossover interaction effects. Psychological Science. D. Darwin. B. Meditation increases compassionate responses to suffering. & Hinton. Ralph Robinson. J. In a social context. Dunne. B. J. G. 2013 . M. Dosenbach. Schlaggar. L. the goals of compassion training and reappraisal training are opposing (otherfocused vs. & Finkel. 11.sagepub.. A.. (in press). & Fischbacher. 646–718).. & McEwen. Weng). R. Davis. (2006). Journal of Personality and Social Psychology. & R. R. these results build on existing evidence that the adult human brain may demonstrate functional and structural changes after mental training (Davidson & McEwen. and Mental Insight Foundation (R. S. Keith and Arlene Bronstein.. Miller. W. 2012. A....sagepub.. & Spinrad. The relationship between training-induced neural changes (DLPFC activation and DLPFC-NAcc connectivity) and altruistic behavior was not unique to compassion trainees. P30-HD003352—M. K. such as psychopaths (Blair. Funding This work was supported by the National Institutes of Health (P01-AT004952 and T32-MH018931—R.. N. J. the Fetzer Institute.. This finding may be due to the differing regulatory goals between compassion training and reappraisal training.. L. Slagter. N. L.. Fehr..). Our findings support the possibility that compassion and altruism can be viewed as trainable skills rather than as stable traits. L. 2. V. Davidson). C.. A unifying view of the basis of social cognition. Cohen. In compassion training. Trends in Cognitive Sciences. (2008). In N. The altruism question. Klingberg. C.g. 396–403. Lutz. 15. & Rizzolatti.html. Keysers. Nature Neuroscience. Eisenberg. 3: Social. See Supplementary Method and Results in the Supplemental Material for detailed method and analyses regarding the independent validation study. B. Hillsdale. 2008) and extend these previous findings to include socioemotional domains such as compassion and altruism. Desbordes. In fact. Cohn.. 387–392. (2004). T. London. NJ: Erlbaum. although a separate validation sample was used to estimate pretraining levels. (2006). University of Wisconsin–Madison Department of Psychology (Hertz Award to This lays the groundwork for future research to explore whether compassion-related trainings can benefit fields that depend on altruism and cooperation (e. (2012). emotional. New York. Future research may build on this study’s findings by measuring altruism at baseline. (2003). Impact at Harvard Libraries on May 22. (Original work published 1871) Davidson. In future research.. 54–58. A multidimensional approach to individual differences in Notes 1. Prosocial development. 1045–1062. Trends in Cognitive Sciences. G. U. 95. A dual-networks architecture of top-down control. whereas the goal in reappraisal training was to decrease personal negative emotions. M. and personality development (pp. P. investigator. JSAS Catalog of Selected Documents in Psychology. 8. U. Furthermore. Y. P. Weng). J. Trends in Cognitive Sciences. 785–791. E. Eisenberg.investigatinghealthyminds.Compassion Training and Altruism 9 also support research suggesting that compassion training enhances positive emotions and neural substrates of affiliation (Klimecki et al. Condon. Vol. Coffey. 12. W. build consequential personal resources. F. principal Downloaded from pss.. (2004). 10.

34.. (2008).. L. 31. Boston. L. J. Lovingkindness: The revolutionary art of happiness. R. 24. J. J. Z. Brain Research Reviews. Retrieved from http://www. . R. Training and plasticity of working memory. J. D.. 48–55. The neuroscience of empathy: Progress. C. C. & Damasio. 26. & Singer. personality. A. (2010). & Singer.. England: Penguin Classics. 88–123. Compassion meditation enhances empathic accuracy and related neural activity. Effect of compassion meditation on neuroendocrine. Neural responses to taxation and voluntary giving reveal motives for charitable donations. 466–469.0017798 Lutz. N.. C. Short-term compassion training increases prosocial behavior in a newly developed prosocial game.1371/journal . & Simon-Thomas. S. J. Adame. Trends in Cognitive Sciences.. T. T. S. E. Hughes.. N. J. & Davidson. Evidence for a frontoparietal control system revealed by intrinsic functional connectivity.. Attention regulation and monitoring in meditation.. (2006).sagepub.0001897 Lutz.. Pace. Empathic neural responses are modulated by the perceived fairness of others. & Cooper. M.. Silani. (2006).. Harbaugh. (2010). Keltner. U. USA.plosone. 8. E. Brown.. An integrative theory of prefrontal cortex function. J. S. 100. A. Psychological Bulletin. & Frith. Lindquist. 54. 59. R. R.. (2011). Nature. & Burghart. (2008).pone. 242–249. & Ochsner. Leiberg. K. T. Slagter. L. Grossman. Kalin. (2001). K.. D.. Current Opinion in Neurology. Lovingkindness meditation increases social connectedness. Regulation of the neural circuitry of emotion by compassion meditation: Effects of meditative expertise. Harris. Cambridge. L. Vincent. S. S.. (2003). and social cognition. Cole... R.. (2010).. 106. Perspectives on Psychological Science. Functional magnetic resonance imaging of reward prediction. .. Morelli. Immordino-Yang. MA: Harvard University Press. Human fronto-mesolimbic networks guide decisions about charitable donation. 4415–4425. . Zaki. PloS ONE. T. & Gross. & Hinton. & Buckner. . Wilson. H. L. 167–202. A. Leiberg. Snyder. . W.. (2008). N. & Grafman. Lovingkindness and compassion meditation: Potential for psychological interventions.. D.. 411–417. .. A.. Zahn. Neuron. J. McColl. T. Journal of Neurophysiology. M. O’Doherty.. E.. G. & Pashler. J. J.. 87–98. P. H. D. 6(3). & Raison.. J. e17798. Compassion: An evolutionary analysis and empirical review. Hutcherson.. I. Cerebral Cortex. 136.. P. Miller. 675–680. Stephan. Unto others: The evolution and psychology of unselfish behavior. T. R. 1126–1132. Rilling. Psychoneuroendocri­ nology..... H.. E. J. T. T. D. F. Raichle. E.. Nature Neuroscience. Moll. K.. G. C. D. 9.. D. M. S... T. Zald. Downloaded from pss. (2008). Schaefer. (1997).. & Wilson. P.pone.. 68.. Dolan. Damasio... H. Neuron. J. T.plosone . D.. T. Wager. 163–169. (2009). Raison. L. e1897. Proceedings of the National Academy of Sciences.. C. (2005). (2011). NeuroImage. & Davidson. (Original work published 1759) Sober. Kahn. (2011). 55.. 2492–2502. NeuroImage.. (2012). K. D. Thurow. B. Puzzlingly high correlations in fMRI studies of emotion. D. J. 103. 317–324.. J. C. Retrieved from http://www. Clinical Psychology Review. K. Klimecki. van Reekum. D. The theory of moral sentiments. (2011).10 Goetz. (1999). Smith. Preuschoff. Prefrontal-subcortical pathways mediating successful emotion regulation. Journal of Neuroscience. W. Decety. (2009). Annual Review of Neuroscience. M. Trends in Cognitive Sciences. 15623–15628. Brefczynski-Lewis. S. B. 14. Ochsner. M.1093/cercor/ bhs142 Klingberg. MA: Shambhala. Hein. R.. P. M. Functional neural plasticity and associated changes in positive affect after compassion training. Krueger.. Social Cognitive and Affective Neuroscience. E. A. 4. D. Hofmann. L... 439. 18.. & Cohen. pitfalls and promise. D. 351–374. 1037–1050. Negi. G. & Singer. & Ochsner. A. Masten. Neural correlates of admiration and compassion. at Harvard Libraries on May 22. L. H.. C. I. Johnstone. Amygdala and ventromedial prefrontal cortex are inversely coupled during regulation of negative affect and predict the diurnal pattern of cortisol secretion among older adults. J. Emo­ tion. R. Mascaro. J. London. (2010). 1622–1625. Lamm. Tenzin Negi. Weng et al. D. M. (2005). Singer. L. Neural responses to ingroup and outgroup members’ suffering predict individual differences in costly helping.. Meta-analytic evidence for common and distinct neural networks associated with directly experienced pain and empathy for pain. Johnstone. 3328–3342. Winkielman. (2013). Mayr. J. Davidson.. C. H. Advance online publication. & Gross. 720–724. Pardini. A. (2007). 8. O. & Eisenberger. 149–160. (2012). L. Klimecki. K. & Singer. innate immune and behavioral responses to psychosocial stress. O. S. Trends in Cognitive Sciences. 8021–8026... C. S. Knutson. Lamm. A. J.. M.. Science.. Batson.. (2006). PLoS ONE.. 3(3). 381–388. J. 12. H. Urry. Salzberg. Davidson. S. 274–290. (2009). (2008). C. 2013 . Proceedings of the National Academy of Sciences. Dunne. N. P. K. The cognitive control of emotion. W. 15..1371/journal. I. E. An fMRI investigation of empathy for “social pain” and subsequent prosocial behavior. P. A. Vul. USA. E. J.. 316. Sivilli. M. Seymour. The human amygdala and the emotional evaluation of sensory stimuli. 41. de Oliveira-Souza. T. T. C. doi:10.. A. L. H.. B.

Sign up to vote on this title
UsefulNot useful